A damning report by the National Audit Office has severely criticised the government's Universal Credit reforms.
Ashwin Kumar is director of Liverpool Economics and a former Senior Civil Servant at the Department for Work and Pensions
Yesterday’s damning report by the National Audit Office (NAO) into Universal Credit was reported as yet another government IT failure: £34 million wasted on unnecessary IT systems with ministers blaming civil servants.
But if you read the NAO’s report, it is clear that these problems were a direct result of the government’s obsession with a completely unrealistic timetable for delivering the changes. If officials had been allowed to use their traditional approach to major benefit changes, national rollout wouldn’t have started until April 2015, too late for Tory election strategists.
Plans for a Universal Credit-style benefit have been long in the making. For many years under Labour, officials were working on plans for what was called a Single Working Age Benefit. The reasoning has remained consistent: sweep away complexity in the benefit system to make work incentives more streamlined and transparent, make the process of moving in and out of work easier for claimants, reduce the potential for error, and reduce delivery costs.
In November 2010, the new government announced plans to bring these ideas to fruition in the form of Universal Credit but insisted upon national rollout from October 2013. The only way to meet this timetable was to abandon the usual method of designing the policy before building the IT systems. The new approach known as ‘agile’ (or ‘extreme project management’) had never before been used by the Department for Work and Pensions (DWP).
The idea was that policy designers were supposed to get feedback from each stage of IT development to help them design the policy. Needless to say, this turned into a shambles. The NAO was blunt: “The source of many problems has been the absence of a detailed view of how Universal Credit is meant to work”.
As the first signs of panic emerged, Iain Duncan Smith announced in November 2011 that the government would be piloting the new benefit in four areas: not a bad idea in itself – Labour policies such as Employment & Support Allowance and the Local Housing Allowance were both piloted before finalising their design.
However Duncan Smith’s new pilots were supposed to start in April 2013, and test the IT systems that were going to be used nationally from October 2013, so there was even less time to deliver the IT changes.
Panic bred panic and, instead of working on the system as a whole, the department concentrated on just getting something to work for the pilots. The only way to do this was to reduce the scope – deliver a ‘pilot-lite’. So a whole raft of security measures were abandoned for the pilots, including online reporting of changes in claimants’ circumstances, the IRIS system for assessing fraud risks, and the IDA system for checking claimants’ identities, and most of the six predecessor benefits are excluded from the pilot.
So right now we have pilots that don’t test what the government says it wants to roll out nationally, no plan for developing all of the components that are missing from the pilots, no clear view of the detail of many aspects of the policy, and no confidence that the government can actually deliver many of the original benefits that were promised.
Iain Duncan Smith has continued to promise that the final rollout of the policy will be completed as originally planned by 2017 and within the originally agreed budget of £2.4 billion. The only way he can deliver this is to make compromise after compromise after compromise in the scope of the project. The question we need to ask ourselves is whether what started out with relatively lofty ambitions to transform the experience of claimants will now be a shadow of its former self.
3 Responses to “Universal Credit shambles uncovered in new report”
jameslaird
Civil servant pen pushers like yourself are totally wrong to try to blame agile processes for this failure. In fact, the DWP set out on the project using the more traditional waterfall approach, only then to turn to agile in the hope of salvaging something from an obviously doomed project. This is a typical governmental failure, nothing more, nothing less.
swatnan
… blame the organ grinder, not the monkeys …
the Quiet Man is getting a bit tetchy these days.
Mason Dixon, Autistic
Except that ‘agile’ has been a buzzword doing the rounds in relation to the UC IT system since at least 2011. Smith said everyone was fine as of this March, not even hinting at any problems.